OBJECTIVES: In France, economic evaluation is used in the process of price negotiation of innovative health products. It’s based on a modelling approach, which requires the choice of Health Economic Decision Models (HEDM). Our objective was to review and analyze the use of HEDM in the cost-effectiveness analyses (CEAs) of innovative health products submitted by the manufacturers.

METHODS: We identified all published economic appraisals of CEAs from January 2016 to December 2022. The main aspects specifying the HEDM were independently extracted by two authors into a predefined grid (e.g., therapeutic area, HEDM type, health states ...). Disagreements were resolved through discussion.

RESULTS: 138 appraisals including HEDM from January 2016 to December 2022 were analyzed. The main therapeutic area was oncology. 30% (42/138) of the CEAs were for orphan drugs. Most (53%) of CEAs were based on Partitioned Survival Model (PSM) while State transition models (STM) such as Markov and semi-Markov models were used in 30% of CEAs. Micro-simulation models were rarely used (4%). Combined models (e.g. decision tree and Markov model) were implemented in 13% of the CEAs. Although an improvement was seen regarding the global quality of CEAs submitted during the studied period, PSMs lacked of documentation on the projection of the survival curves even if assumed to be adequate to reflect the disease history. The use of STM was fairly justified. Whatever the type of HEDM, the assumptions of the modelling structure were subject to important limitations in almost of a quarter of the CEAs.

CONCLUSIONS: This review showed that PSM were frequently used. In depth documentation and adequate exploration of structural uncertainty related to the model assumptions should be provided.